While overexpression of some cancer-specific or cancer-associated genes could contribute to the onset and progression of human cancer, malignant tumor cells could also inappropriately express and regulate a preexisting normal cell program, leading to these proteins to be perpetually activated or unrestrained in malignant cells compared with normal counterparts. In this proposal, we present the case of a closely controlled protease (matriptase) at interepithelial junctions that has been constitutively activated and inappropriately distributed to the invading fronts of cancer cells. Matriptase is broadly expressed by almost all human epithelial tissues, suggesting that the physiological role of matriptase may be associated with some rudimentary feature of epithelium, such as interepithelial junctions. Indeed, we have observed that in nontransformed mammary epithelial cells activation of matriptase in response to its physiological, blood- borne activator sphingosine 1-phosphate (S1P) only occurs on intercellular junctions. Furthermore, atypical protein kinase C zeta, a tight junction protein, is likely to be involved in $1P-indcued matriptase activation. In contrast, breast cancer cells constitutively activate matriptase regardless of the presence of S1P, and the activated matriptase is not restricted to cell-cell contacts and has been detected on membrane ruffles and within the cells. Therefore, in breast cancer cells matriptase could serve as membrane activator to recruit and activate urokinase-type plasminogen activator (uPA) and hepatocyte growth factor (HGF), an extracellular matrix-degrading protease system and cell growth/motility factor, respectively. In the current grant proposal, we will carry out four aims. First, we will elucidate physiological role of matriptase by investigating its functional localization and searching for its physiologically relevant substrates and target genes. Second, we will investigate the molecular mechanisms whereby matriptase activity is closely regulated. Third, we will characterize binding proteins of matriptase from matriptase complexes. Finally, we will investigate how matriptase activity, in a model system which matriptase activation can be enhanced or suppressed, affects the phenotypes of breast cancer cells. These studies could lead to new perspectives on the deregulation of this physiological protease in breast cancer and provide new avenues for diagnosis and intervention of the disease.